id stringlengths 36 36 | source stringclasses 15
values | formatted_source stringclasses 13
values | text stringlengths 2 7.55M |
|---|---|---|---|
5346f604-cd95-4a9b-8c94-4850869d24fa | trentmkelly/LessWrong-43k | LessWrong | Exploring out-of-context reasoning (OOCR) fine-tuning in LLMs to increase test-phase awareness
Summary
The Stages-Oversight benchmark from the Situational Awareness Dataset tests an LLM’s ability to distinguish evaluation prompts from deployment prompts. This capability is a prerequisite for many rogue LLM threat scen... |
f48420e3-fe41-45b6-a002-62d7726bcea4 | trentmkelly/LessWrong-43k | LessWrong | Rationalist Gnosticism
Gnosticism was an old Christian sect which solved the problem of evil by asserting that the material world was created by an evil entity, and that only inner enlightment can reveal the path to the true, good and loving god. I think many rationalists have a worldview similar to this, perhaps most... |
9b78585e-706c-43b6-a073-f515124cfe81 | trentmkelly/LessWrong-43k | LessWrong | Communication and Waking Hours
One of the ways technology interacts with culture is about who's job it is to avoid contacting people at bad times. With older technologies there were clear conventions based on their limitations:
* Mail: you can send the letter whenever, and the person will deal with it when convenient... |
8724758a-4267-4cc5-b595-aacd17f1a46e | trentmkelly/LessWrong-43k | LessWrong | Gamification and rationality training
Watch this Extra Credits and take note especially of the end, where they're offering consultation to educators and doctors re making educational stuff and other useful stuff more like games. (Incidentally, the people that make Extra Credits all work in or have been around in the g... |
8ca9757b-24c2-4189-bdf0-7cc744b46b87 | trentmkelly/LessWrong-43k | LessWrong | How I Learned to Stop Worrying and Love MUM
Epistemic status: Speculative
The title is a tongue-in-cheek reference to Google AI's latest showcase: Multitask Unified Model, or MUM for short. Further details can be found in their arXiv paper; Rethinking Search: Making Experts out of Dilettantes.
Let's say hi to MUM.
... |
3832baaa-ccb7-4eaf-a1e3-c12a3e9ad142 | trentmkelly/LessWrong-43k | LessWrong | Game Theory & The Golden Rule (From Reddit)
|
7b7beae8-dcc8-4cd0-bf72-8d0d7af63031 | trentmkelly/LessWrong-43k | LessWrong | How counting neutrons explains nuclear waste
You probably recall from high school chemistry that atoms are made up of a nucleus containing protons and neutrons, surrounded by electrons. But how many of each?
If you remember a little bit more from high school chemistry, you’ll recall that the number of protons determi... |
3e08762c-6ac4-4173-a89a-a7e5fcb118d6 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | How to Keep Improving When You're Better Than Any Teacher - Iterated Distillation and Amplification
hi today we're going to talk about
iterated distillation and amplification
so let's say we want to play go and we
want to be really good at it so we're
trying to create a function which if
we're given a go board in a pa... |
803c888d-cd22-41e3-b102-3243f5f0ce81 | trentmkelly/LessWrong-43k | LessWrong | What Should AI Owe To Us?
Accountable and Aligned AI Systems via Contractualist AI Alignment
This is an extended and edited transcript of the talk I recently gave at EAGxSingapore 2022.
The title has been changed for easier searchability of "Contractualist AI Alignment".
Abstract: Artificial intelligence seems poise... |
d0a5c351-c464-4009-8026-63b5b08ff925 | StampyAI/alignment-research-dataset/arxiv | Arxiv | A Base Camp for Scaling AI
1. Introduction and Related Work
----------------------------------
AI has had a long and difficult childhood [[4](#bib.bib4)], with nine “AI Winters” of
varying degrees commonly identified [[31](#bib.bib31)]. A key technical roadblock seems
to be the problem of scale, both in terms of ... |
a1dc3296-4d4a-4b16-a4e2-a58692fa506f | trentmkelly/LessWrong-43k | LessWrong | Cataloging my skills - good idea?
tl,dr: Discuss utility of writing down a portfolio of bayesian evidence for/against your skills, and keeping it for personal reference.
Last night, I applied for the Write/Design Rationality Curriculum Position. I am the sort who sincerely needs the continuous requests to err on the ... |
073a9a11-828d-4409-a083-934a2b100116 | trentmkelly/LessWrong-43k | LessWrong | How to make plans?
What models are out there for how humans can personally make plans, especially grand master plans that scale to a whole lifespan or beyond?
* What is the set of good plan components? (e.g. "backchaining")
* What useful tools are out there to help with planning? (e.g. "pen and paper")
* How does ... |
bd767449-42d6-4e7d-9ad5-28954d624467 | trentmkelly/LessWrong-43k | LessWrong | Reflecting on the transhumanist rebuttal to AI existential risk and critique of our debate methodologies and misuse of statistics
This morning, I read an article that counter-argues many of our arguments about AI existential risk, and I'm curious what other people's thoughts are on the arguments, linked here: https://... |
9a175f60-f951-46ac-9cf1-0645e67d102c | trentmkelly/LessWrong-43k | LessWrong | Community roles: teachers and auxiliaries
Related to: Building rationalist communities, Lessons from Latter-day Saints, Holy Books (Or Rationalist Sequences) Don't Implement Themselves, Designing rationalist projects.
I'm beginning a new subseries of posts, trying to answer the following question: what should be the ... |
d7e53326-ef8e-43ce-a60e-c81cacf3ae10 | StampyAI/alignment-research-dataset/youtube | Youtube Transcripts | Channel Introduction
hi my name is Rob Myles welcome to my
channel
it seems pretty likely that sooner or
later will develop general artificial
intelligence that is to say a software
system that's able to reason about the
world in general and take actions in the
world at large we don't know how to do
that yet
but befor... |
c35e7d4e-86a3-43d2-a8b8-657117acb06d | StampyAI/alignment-research-dataset/blogs | Blogs | A central AI alignment problem: capabilities generalization, and the sharp left turn
(*This post was factored out of a larger post that I (Nate Soares) wrote, with help from Rob Bensinger, who also rearranged some pieces and added some text to smooth things out. I’m not terribly happy with it, but am posting it anyway... |
d99bd4d3-a7a7-48b1-aed4-bbeb14423c53 | trentmkelly/LessWrong-43k | LessWrong | Hate Crime Statistics
It's generally hard to tell whether long-term problems are getting better or worse from looking at news coverage or hearing about problems. Each gives you some information, but it's hard to tell how it compares to what was happening before. Quantitative data can help give some perspective, howeve... |
c6cf25c3-3db5-48da-be10-938967a6f733 | trentmkelly/LessWrong-43k | LessWrong | Effective Empathy
I have just summarised a video by Psychiatrist Alok Kanija for myself, which is about maximising the benefits of empathy and lowering the drawbacks. It generally pleads the case, that people are getting more selfish and naricicistic, but the rarer empathy becomes the bigger the benefits become, sort ... |
d81a604e-f8b1-484b-aaaf-d26dad6f3c91 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Israel Less Wrong Meetup - Social, Board Games
Discussion article for the meetup : Israel Less Wrong Meetup - Social, Board Games
WHEN: 11 November 2014 07:00:00PM (+0200)
WHERE: Yigal Alon Street 98, Tel Aviv-Yafo, Israel
We're going to have a meetup on Thursday, August 7th at Google Israel's offices, Ele... |
a97fecb9-7553-4d57-8498-fad632c076ac | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Announcing AISIC 2022 - the AI Safety Israel Conference, October 19-20
ALTER, with the University of Haifa and the Technion, is excited to announce the details about its upcoming conference introducing AI safety in Israel. We’re going to be hosting Stuart Russell, as well as several other AI safety researchers, to spe... |
8ba694c2-cf29-494d-a169-79f3a528b2f3 | trentmkelly/LessWrong-43k | LessWrong | EAI Alignment Speaker Series #1: Challenges for Safe & Beneficial Brain-Like Artificial General Intelligence with Steve Byrnes
A couple months ago EleutherAI started an alignment speaker series, some of these talks have been recorded. This is the first instalment in the series. The following is a transcript generated ... |
2eb87e95-a91a-41dd-9553-917b6d5b6577 | StampyAI/alignment-research-dataset/blogs | Blogs | psi: a universal format for structured information
psi: a universal format for structured information
--------------------------------------------------
representing structured information in general is not a solved problem.
in this post i outline a format tentatively called "psi", proposing what i think is the be... |
5782a5a4-8145-407a-b1b0-9cb52fb4747f | trentmkelly/LessWrong-43k | LessWrong | Random personal thoughts on workload pessimism as a procrastination problem
I am in one of those moods where I feel extremely burned out and apathetic. What I am noticing about my feelings right now is that there is no short term reward system that could possibly motivate me. I don't want any candy, food, soda, coffee... |
3a1195eb-097f-4fd2-9780-441a8b9d0c5f | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | What are some non-purely-sampling ways to do deep RL?
.mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: normal; white-space... |
6c047e1a-f113-4b9f-a13f-cea2ab2f33ca | trentmkelly/LessWrong-43k | LessWrong | Multidimensional signaling
What do you infer about a person who has ugly clothing? Probably that they have poor taste (in clothes, or subcultures). But it could also be that they are too poor to improve their wardrobe. Or can’t be bothered.
What about someone with poor grades? The obvious inference is that they aren... |
6d3844ae-b4bf-431e-9d94-149461d02697 | trentmkelly/LessWrong-43k | LessWrong | Possible AI “Fire Alarms”
Eliezer says that there is no fire alarm for AGI and I agree that there’s no single event that would wake everyone up. On the other hand, there are a lot of possible events that could cause a significant shift in the Overton window or wake a lot of people. Since this seems strategically impor... |
d564ea8e-2cc9-4469-8984-5c40e028a7f6 | StampyAI/alignment-research-dataset/special_docs | Other | Wireheading Done Right: Stay Positive Without Going Insane
Wireheading Done Right
Wireheading Done Right: Stay Positive Without Going
Insane
Andr es G omez Emilsson
Qualia Research Institute
San Francisco, CA
algekalipso@gmail.com
Abstract
Wireheads are beings who have changed their reward architecture in order to b... |
daaeb179-c229-473c-bd76-1bd150a5868f | trentmkelly/LessWrong-43k | LessWrong | Investigating the Ability of LLMs to Recognize Their Own Writing
This post is an interim progress report on work being conducted as part of Berkeley's Supervised Program for Alignment Research (SPAR).
Summary of Key Points
* We test the robustness of an open-source LLM’s (Llama3-8b) ability to recognize its own out... |
f34d5965-4863-4128-b9db-df2b458fc7d8 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Tel Aviv Meetup
Discussion article for the meetup : Tel Aviv Meetup
WHEN: 27 January 2015 07:00:00PM (+0200)
WHERE: Google, Electra Tower 29th Floor, Yigal Alon Street 98, Tel Aviv.
This meeting we're going to try something a little different. Instead of having a formal lecture given by one of us we're goi... |
f7f7b5b1-43f9-45bf-8220-a0345cecc312 | trentmkelly/LessWrong-43k | LessWrong | Aligning AI by optimizing for "wisdom"
In this post, we’ll introduce wisdom as a measure of the benevolence and internal coherence of an arbitrary agent. We’ll define several factors, such as the agent’s values, plans, evidence, and alignment with human values, and then define wisdom as consistency within and between ... |
2acada2e-bbe8-4208-b322-6a3a91199e37 | trentmkelly/LessWrong-43k | LessWrong | Prioritizing the Arts in response to AI automation
In modern society, many find meaning and fulfillment from their work and careers. I, for one, certainly get enjoyment from a productive day's work.
As a software engineer, I've benefited greatly from GPT-3. I estimate that I'm 5-10 times more productive on a large ra... |
9e72c670-db9b-4f3d-acd1-97e611116c5e | StampyAI/alignment-research-dataset/special_docs | Other | AI Alignment Research Overview
---
identifier: 045a6d13-a46c-4743-aaf8-cd1490df6e98
language: en
title: AI Alignment Research Overview \[Oct 2019\]
---
[]{#AIAlignmentResearchOverviewOct2019.xhtml}
[AI Alignment Research Overview]{.c15} {#AIAlignmentResearchOverviewOct2019.xhtml#h.n1wk9bxo847o .c28}
----------------... |
937be38e-3242-45c2-b507-dacb7324ecc9 | trentmkelly/LessWrong-43k | LessWrong | Why I Reject the Correspondence Theory of Truth
This post began life as a comment responding to Peer Gynt's request for a steelman of non-correspondence views of truth. It ended up being far too long for a comment, so I've decided to make it a separate post. However, it might have the rambly quality of a long comment ... |
f69e8086-1369-4991-bb4a-79462b575393 | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Implications of large language model diffusion for AI governance
*This post is one part of the sequence* [*Understanding the diffusion of large language models*](https://forum.effectivealtruism.org/s/8rYkpiFhbb4HsbzFc)*. As context for this post, I strongly recommend reading at least the* [*5-minute summary*](https:/... |
e8d77a16-e1cb-4bb9-a667-246e978d7152 | trentmkelly/LessWrong-43k | LessWrong | Utilitarianism and the replaceability of desires and attachments
Summary
1. Consider a pill that would cause a happy person with a fulfilling life to abandon their most important desires and cherished attachments, including goals, career and loved ones, but increase their lifetime subjective well-being. If what’s bes... |
b841fc14-0d48-46d3-9a5d-57b09a44fc00 | trentmkelly/LessWrong-43k | LessWrong | Wildfire of strategicness
It may not be feasible to make a mind that makes achievable many difficult goals in diverse domains, without the mind also itself having large and increasing effects on the world. That is, it may not be feasible to make a system that strongly possibilizes without strongly actualizing.
But su... |
b06d21fb-75fb-45a3-8071-dd503fb25948 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Arguing all sides with ChatGPT
The practice of seeing what ChatGPT has to say on an issue is popular, but as ChatGPT itself says, "As an AI language model, I do not have beliefs or biases, and I can provide arguments for any side of an issue."
The following is a complete, unedited conversation I have just had with C... |
93c9cf75-35ed-49a5-9ee3-951c6bb90377 | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] Engelbart: Insufficiently Recursive
Today's post, Engelbart: Insufficiently Recursive was originally published on 26 November 2008. A summary (taken from the LW wiki):
> The development of the mouse did lead to a productivity increase. But it didn't lead to a major productivity increase at creating fut... |
8ea16344-2bfa-4bb7-bc6a-d87dd3203f6b | StampyAI/alignment-research-dataset/blogs | Blogs | June 2016 Newsletter
| |
| --- |
|
**Research updates**
* New paper: “[Safely Interruptible Agents](https://intelligence.org/2016/06/01/new-paper-safely-interruptible-agents/).” The paper will be presented at UAI-16, and is a collaboration between Laurent Orseau of Google DeepMind and Stuart Armstrong of the Future... |
c2d2a042-b6f8-4ad0-bcfc-b8d31b9da4a6 | trentmkelly/LessWrong-43k | LessWrong | Fast Minds and Slow Computers
The long term future may be absurd and difficult to predict in particulars, but much can happen in the short term.
Engineering itself is the practice of focused short term prediction; optimizing some small subset of future pattern-space for fun and profit.
Let us then engage in a bit of... |
5bf44890-d4bc-427b-acc7-da387e013d5a | trentmkelly/LessWrong-43k | LessWrong | Kevin Simler: Outbreak
Kevin Simler released another blogpost, this time on disease spread modeling. Excerpt:
> Harry Stevens at The Washington Post recently published a very elegant simulation of how a disease like COVID-19 spreads. If you haven't already, I highly recommend checking it out.
>
> Today I want to fo... |
04cc8250-877e-4685-a99e-db58dd6e32b4 | trentmkelly/LessWrong-43k | LessWrong | State Space of X-Risk Trajectories
Justin Shovelain developed the core ideas in this article and assisted in writing, David Kristoffersson was the lead writer and editor.
Cross-posted on the EA Forum
Abstract
Currently, people tend to use many key concepts informally when reasoning and forming strategies and polici... |
41b3bf83-f065-47f1-874d-fd37874657a9 | trentmkelly/LessWrong-43k | LessWrong | 9 Strategies for Effective Donors
Summary: This piece is mainly of relevance to EA-oriented Less Wrongers. It provides 9 strategies meant to help donors be more effective. It's prompted by conversations about the desire for more introductory-style pieces on various issues aimed at newer members of the EA movement who... |
9190745e-f50a-4d54-95b2-aa313b75f39c | trentmkelly/LessWrong-43k | LessWrong | [Link] Epigenetics
Your daily dose of science knowledge will once more be provided by Gregory Cochran, clearing up some misconceptions you may have heard about Epigenetics.
> As I understand it, in some circles, there is a burgeoning hope that practice in this generation will somehow improve performance in the next ... |
f6be9e80-52a0-4d21-bcab-e0d5f9e499c4 | trentmkelly/LessWrong-43k | LessWrong | Summary and Lessons from "On Combat"
On Combat - The Psychology and hysiology of Deadly Conflict in War and in Peace by Lt. Col. Dave Grossman and Loren W. Christensen (third edition from 2007) is a well-written, evidence-based book about the reality of human behaviour in life-threatening situations. It is comprehens... |
dcb31e46-aef6-45b2-9d67-f3ae3757141f | trentmkelly/LessWrong-43k | LessWrong | Meetup : Frankfurt
Discussion article for the meetup : Frankfurt
WHEN: 23 March 2014 02:00:00PM (+0100)
WHERE: Frankfurt
Location will be published in time. Contact me under 0176 34 095 760. I prefer texting to calling. If you have any special requirements and need help to attend (whether a disability, social anxie... |
4fbf29b9-ffd4-4da0-a6ef-20bb9de073d7 | StampyAI/alignment-research-dataset/lesswrong | LessWrong | Personal imitation software
The question is very simple: could you train a powerful computer program to act like an accurate copy of yourself?
Some LW members may have already tried this, but it would require super powerful software. Even if it's only text based, GPT-3 wouldn't be enough. Maybe GPT-13.
If such a p... |
dc96abb7-2795-4b30-a032-c8c6e7b22b5b | trentmkelly/LessWrong-43k | LessWrong | What is Inadequate about Bayesianism for AI Alignment: Motivating Infra-Bayesianism
Introduction
Infra-Bayesianism is a mathematical framework for studying artificial learning and intelligence that developed from Vanessa Kosoy’s Learning Theoretic AI Alignment Research Agenda. As applied to reinforcement learning, th... |
b4114440-1e9e-48fc-878b-6f8de4ed3e30 | trentmkelly/LessWrong-43k | LessWrong | Do we Share a Definition for the word "ideal"?
Related: A Proposal for a Simpler Solution To All These Difficult Observations and Problems
Perhaps this will be seen as spam and that I have broken a rule of propriety, but my previous discussion has already served its purpose and I think will continue to do so. I need... |
453b3642-51bc-4c96-aa21-970d3117e0fe | trentmkelly/LessWrong-43k | LessWrong | A majority coalition can lose a symmetric zero-sum game
Just a neat little result I found when thinking about Jessica's recent post.
For n players, let ai be the action of player i and si(a1,a2,...an) the reward of player i as a function of the actions of all players. Then the game is symmetric if for any permutation... |
e8485b68-9d93-480d-b6f3-b34a4796bb4d | StampyAI/alignment-research-dataset/lesswrong | LessWrong | The Lifespan Dilemma
One of our most controversial posts ever was "[Torture vs. Dust Specks](/lw/kn/torture_vs_dust_specks/)". Though I can't seem to find the reference, one of the more interesting uses of this dilemma was by a professor whose student said "I'm a utilitarian consequentialist", and the professor said ... |
df6180a7-ac98-4758-b53e-8001234099d5 | trentmkelly/LessWrong-43k | LessWrong | David Burns Thinks Psychotherapy Is a Learnable Skill. Git Gud.
Epistemic status: I listened to this, and I found these statements surprising in a “big if true” way. It didn't seem obviously false, but I am very much relying on him here. If someone thinks his claims are completely made up, please let me know in the co... |
ba913f37-d572-4c4f-ad52-cf3fb4a38473 | StampyAI/alignment-research-dataset/blogs | Blogs | Outcomes of inducement prizes
*Published 29 August 2022; last updated 30 August 2022*
*This project is incomplete, contains information that may change over time, and may or may not be updated in the future*
This is a dataset of prizes we could find for incentivizing progress toward a specific technical or intelle... |
ffd26295-1d1e-4919-a499-10c80aea0c93 | trentmkelly/LessWrong-43k | LessWrong | [Poll] Method of Recruitment
In another thread, we have been discussing how people (especially female people) have come to find out about LessWrong. Instead of just guessing, I figured I would make a poll.
I remember in recent history there was a thread on the subject, but the answers were mainly "I got here from HPM... |
f444c724-c0a6-4399-885d-f1678eaf7792 | trentmkelly/LessWrong-43k | LessWrong | Meetup : Schelling Cafe
Discussion article for the meetup : Schelling Cafe
WHEN: 29 August 2015 03:00:57PM (-0700)
WHERE: Cafe La Renaissance, 3045 Shattuck Ave, Berkeley, CA 94705
For anyone in the Berkeley, SF, Oakland area, there's a weekly gathering called the Schelling Cafe that holds on Saturdays at 3pm (on h... |
03395a2e-8bdc-44ba-8a9c-24d42188c46e | trentmkelly/LessWrong-43k | LessWrong | The mathematical universe: the map that is the territory
This post is for people who are not familiar with the Level IV Multiverse/Ultimate Ensemble/Mathematical Universe Hypothesis, people who are not convinced that there’s any reason to believe it, and people to whom it appears believable or useful but not satisfact... |
1a164c8a-e129-4adc-b1fc-635bb1d54e01 | trentmkelly/LessWrong-43k | LessWrong | 2019 Review Rewrite: Seeking Power is Often Robustly Instrumental in MDPs
For the 2019 LessWrong review, I've completely rewritten my post Seeking Power is Often Robustly Instrumental in MDPs. The post explains the key insights of my theorems on power-seeking and instrumental convergence / robust instrumentality. The ... |
8cc1ead6-27f1-4e86-a80f-faeea8df83fc | trentmkelly/LessWrong-43k | LessWrong | Non-Book Review: Patterns of Conflict
Soviet MiG-15 (left) and American F-86 (right)
From 1951-53, the Korean war saw the first major dogfights between jet aircraft - mainly the American F-86 and the Soviet MiG-15. On paper, the MiG-15 looked like the superior aircraft: it could out-climb, out-accelerate, out-turn, an... |
0100aa00-f4d4-48c1-89da-e47151d3a948 | trentmkelly/LessWrong-43k | LessWrong | Could we breed/engineer intelligent parrots?
Advantages:
* Breathes air (unlike octopi)
* Talks (unlike octopi)
* No IVF surgeries required (unlike mammals)
* Very long lifespan — time to learn more
* Highly social — good for communication
* 50 million domestic parrots worldwide — enough to find strong correlat... |
f6cca954-fc9f-4fb8-8246-96a76b551845 | trentmkelly/LessWrong-43k | LessWrong | Shannon Vallor’s “technomoral virtues”
Shannon Vallor is the first Baillie Gifford Chair in the Ethics of Data and Artificial Intelligence at the Edinburgh Futures Institute. She is the author of Technology and the Virtues: A Philosophical Guide to a Future Worth Wanting. She believes that we need to build and cultiva... |
f872015b-35dd-40ef-8373-b4d1c29cf9cb | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | Sydney AI Safety Fellowship
I’m excited to announce the launch of the Sydney AI Safety Fellowship which will provide fellows from Australia and New Zealand the opportunity to pursue projects in AI Safety or spend time upskilling. These projects may be technical projects, projects related to policy, or movement buildin... |
ef5725e8-f366-46a1-8b72-e345b7a8c00f | trentmkelly/LessWrong-43k | LessWrong | Human Minds are Fragile
We are familiar with the thesis that Value is Fragile. This is why we are researching how to impart values to an AGI.
Embedded Minds are Fragile
Besides values, it may be worth remembering that human minds too are very fragile.
A little magnetic tampering with your amygdalas, and suddenly y... |
81bbdd1f-96bf-46ba-ac07-d4d01982e184 | trentmkelly/LessWrong-43k | LessWrong | Fake Morality
God, say the religious fundamentalists, is the source of all morality; there can be no morality without a Judge who rewards and punishes. If we did not fear hell and yearn for heaven, then what would stop people from murdering each other left and right?
Suppose Omega makes a credible threat that if you... |
5c53ada8-db5e-4bbd-9bed-28d2211f46e6 | trentmkelly/LessWrong-43k | LessWrong | A gene for bad memory? (Link)
Original paper: "RGS14 is a natural suppressor of both synaptic plasticity in CA2 neurons and hippocampal-based learning and memory"
Sci. Am. article: "Knocking Out a 'Dumb' Gene Boosts Memory in Mice"
Science Daily article: "Gene Limits Learning and Memory in Mice"
They haven't found ... |
b8915b01-efe8-42e1-8584-c557ae6423db | trentmkelly/LessWrong-43k | LessWrong | Conspicuous saving
It's well known that people around the world can be really bad at saving money. Pre-pandemic, Americans saved an average of 7.6% of their income, electing to spend the rest. If people saved more money and invested it, then they'd be able to retire earlier, be wealthier, and therefore probably be bet... |
bae57611-e2ea-4b9f-9914-8c4592704608 | trentmkelly/LessWrong-43k | LessWrong | Philosophical love poetry
Because Daily Nous asked for it.
***
Roses are red
And their redness canonic
I’m longing to take you
To realms less platonic
***
Violets are bled
Roses are rue
Induction’s mysterious
But long live ye and mou
***
Roses are red,
And I’d read rose reviews
But I viewed your re... |
8b0f7cb1-8e2d-4cb1-9a0f-183f3bbd704a | trentmkelly/LessWrong-43k | LessWrong | Representation Engineering has Its Problems, but None Seem Unsolvable
TL;DR: Representation engineering is a promising area of research with high potential for bringing answers to key challenges of modern AI development and AI safety. We understand it is tough to navigate it and urge all ML researchers to have a close... |
7358a9ff-de12-47ad-87e2-88ea8a9a4706 | trentmkelly/LessWrong-43k | LessWrong | Work on Security Instead of Friendliness?
> So I submit the only useful questions we can ask are not about AGI, "goals", and other such anthropomorphic, infeasible, irrelevant, and/or hopelessly vague ideas. We can only usefully ask computer security questions. For example some researchers I know believe we can achiev... |
75b79867-d2aa-4b20-a9bc-1e4f6cae9842 | trentmkelly/LessWrong-43k | LessWrong | Absolute Authority
The one comes to you and loftily says: “Science doesn’t really know anything. All you have are theories—you can’t know for certain that you’re right. You scientists changed your minds about how gravity works—who’s to say that tomorrow you won’t change your minds about evolution?”
Behold the abyssal... |
3a937aba-fae2-47fb-b2f3-5bd67998fa3d | trentmkelly/LessWrong-43k | LessWrong | Polio Lab Leak Caught with Wastewater Sampling
I hadn't seen this discussed here:
> The near complete eradication of wildtype polioviruses (WPV) means that strict containment by facilities for essential work with infectious WPV is required. In the Netherlands, we have implemented environmental surveillance around all... |
441fc0bb-c6e3-4040-b894-78ad123325d6 | trentmkelly/LessWrong-43k | LessWrong | A brief review of the reasons multi-objective RL could be important in AI Safety Research
By Ben Smith, Roland Pihlakas, and Robert Klassert
Thanks to Linda Linsefors, Alex Turner, Richard Ngo, Peter Vamplew, JJ Hepburn, Tan Zhi-Xuan, Remmelt Ellen, Kaj Sotala, Koen Holtman, and Søren Elverlin for their time and kind... |
cfaca13e-7703-4634-9d76-4c9b312eded8 | trentmkelly/LessWrong-43k | LessWrong | Utopia for artificial minds
TLDR: Alignment as we currently define it seems hard because we recognize that what we as humans want is pretty arbitrary from a non-human perspective. In contrast, global utility maximization is something that an ASI might independently discover as a worthwhile goal, regardless of any huma... |
c9833c72-a2ac-4b04-8c29-b297427eb90f | trentmkelly/LessWrong-43k | LessWrong | [SEQ RERUN] In Praise of Boredom
Today's post, In Praise of Boredom was originally published on 18 January 2009. A summary (taken from the LW wiki):
> "Boredom" is an immensely subtle and important aspect of human values, nowhere near as straightforward as it sounds to a human. We don't want to get bored with brea... |
38ee9bd4-0bd9-4bda-8458-e17add898484 | trentmkelly/LessWrong-43k | LessWrong | Tell Culture
Followup to: Ask and Guess
Ask culture: "I'll be in town this weekend for a business trip. Is it cool if I crash at your place?" Response: “Yes“ or “no”.
Guess culture: "Hey, great news! I'll be in town this weekend for a business trip!" Response: Infer that they might be telling you this because they w... |
e52d6895-5bc4-4a0b-84c3-4826d6367ec3 | trentmkelly/LessWrong-43k | LessWrong | [Link] OpenAI: Learning to Play Minecraft with Video PreTraining (VPT)
https://openai.com/blog/vpt/
> We trained a neural network to play Minecraft by Video PreTraining (VPT) on a massive unlabeled video dataset of human Minecraft play, while using only a small amount of labeled contractor data. With fine-tuning, our... |
28f60bc5-408d-4d44-96b3-ff62acca2cd5 | trentmkelly/LessWrong-43k | LessWrong | [Link] Reason: the God that fails, but we keep socially promoting….
An interesting blog post by Razib Khan, who many here probably know from his Gene Expression blog, the old gnxp site or perhaps from his BHTV debate with Eliezer.
> One point which I’ve made on this weblog several times is that on a whole range of is... |
7cacbf2e-0a4e-435c-b420-49790cd5b512 | trentmkelly/LessWrong-43k | LessWrong | College advice for people who are exactly like me
It’s college decision season! To celebrate, I’ve been thinking about what I would have told myself in 2011 when I was deciding where to attend. Here’s a stitching-together of several emails I sent to friends who asked for college advice, plus a few additional thoughts.... |
4af17136-ddfb-40b7-a1d2-28d1989ec646 | StampyAI/alignment-research-dataset/arxiv | Arxiv | The Relevance of Proofs of the Rationality of Probability Theory to Automated Reasoning and Cognitive Models
1
The Relevance of Proofs of the Rati onality of Probability Theory to
Automated Reasoning and Cognitive Models
Ernest Davis, Computer Science Department, New York University
Abstract
A number of well-kno... |
dfeb71cd-b603-4772-8926-78f505c0af61 | trentmkelly/LessWrong-43k | LessWrong | Adversarial (SEO) GPT training data?
I guarantee someone's thinking about this, but I haven't seen the scammers selling it yet, so I don't know how transparent or discoverable the scraping/input methods are for LLM source data.
Is there any indication that websites or publishers are modifying their pages/data in ways... |
464b3320-a748-462b-93a6-a37eef8a380d | trentmkelly/LessWrong-43k | LessWrong | The greatest good for the greatest number - starting soonest, or ending last, or lasting longest?
The first greatest good for the greatest number for the greatest number will start "first" (by whatever measurement is applied) but ends before the second greatest good ends and doesn't last as long (in total) as the thir... |
35bb628d-ac98-4e0f-8b05-27325fde1946 | trentmkelly/LessWrong-43k | LessWrong | Clip keys together with tiny carabiners
I've always found it annoying to have a bunch of keys in my pocket (they're heavy, sometimes poke into you, and create weird bulges), so I try to carry as few as possible. Unfortunately, I've reached a point where I have a fairly large number of keys that need to keep with me so... |
d8f7965d-eb91-49eb-916f-f18d3daa65f6 | StampyAI/alignment-research-dataset/blogs | Blogs | Time for AI to cross the human range in StarCraft
*Published 20 Oct 2020; updated 22 Oct 2020*
Progress in AI StarCraft performance took:
* ~0 years to reach the level of an untrained human
* ~21 years to pass from beginner level to high professional human level
* ~2 years to continue from trained human to current... |
52358481-3b75-4610-b9bc-2a3b92ff7220 | trentmkelly/LessWrong-43k | LessWrong | Acquiring common human skills
I used to sometimes pay other people to cut up my meat into small pieces (not as weird as it sounds—our house has a tiny internal token economy). Somehow chewing meat was surprisingly hard for me, and so (unrelatedly?) was cutting it with a knife. I think I attributed these failings to so... |
742c2495-8716-4a48-9fec-513976d30c07 | trentmkelly/LessWrong-43k | LessWrong | The case for "mental strength"
Recently, I was thinking about the differences between physical health and physical fitness, and wondering if there is an analogous distinction on the mental side. Googling, I found that the term "mental strength" is often used for the mental counterpart to physical fitness; "mental fitn... |
4b252cb3-ddc0-453e-9778-273eee3f9e83 | trentmkelly/LessWrong-43k | LessWrong | European Community Weekend 2015 - Followup
Seven months ago, the Berlin LW community announced the second European LessWrong Community Weekend. We wrote:
> From June 12th to 14th awesome people from all across Europe are coming to Berlin to meet, exchange ideas and start projects. The focus is on forming new and stre... |
e4888ce3-5b8a-425b-9eec-a932ddeba7ac | StampyAI/alignment-research-dataset/eaforum | Effective Altruism Forum | Apply now for the EU Tech Policy Fellowship 2023
**Announcing the**[**EU Tech Policy Fellowship 2023,**](https://www.trainingforgood.com/europe-tech-policy) **an 8-month programme to catapult ambitious graduates into high-impact career paths in EU policy, mainly working on the topic of AI Governance.**
Summary
====... |
37338d92-584e-4e10-84c8-08714c4087c6 | trentmkelly/LessWrong-43k | LessWrong | o3 Is a Lying Liar
I love o3. I’m using it for most of my queries now.
But that damn model is a lying liar. Who lies.
This post covers that fact, and some related questions.
O3 IS A LYING LIAR
The biggest thing to love about o3 is it just does things. You don’t need complex or multi-step prompting, ask and it wil... |
1fb5eeb3-961f-4163-a2b6-a9fb0e7efbb4 | StampyAI/alignment-research-dataset/alignmentforum | Alignment Forum | AI Neorealism: a threat model & success criterion for existential safety
Threat Model
============
There are *many* ways for AI systems to cause a catastrophe from which Earth-originating life could never recover. All of the following seem plausible to me:
* **Misuse**: An AI system could help a human or group of hu... |
d21e30ff-e80e-4337-b0a7-2a97c46c6237 | trentmkelly/LessWrong-43k | LessWrong | Forecasting Newsletter: July 2021
Highlights
* Biatob is a new site to embed betting odds into one’s writing
* Kalshi, a CFTC-regulated prediction market, launches in the US
* Malta in trouble over betting and gambling fraud
Index
* Prediction Markets & Forecasting Platforms
* In The News
* Blog Posts
* Long ... |
e8448966-3de5-4cec-8cab-888183195001 | trentmkelly/LessWrong-43k | LessWrong | Rationality Quotes Thread June 2015
Another month, another rationality quotes thread. The rules are:
* Please post all quotes separately, so that they can be upvoted or downvoted separately. (If they are strongly related, reply to your own comments. If strongly ordered, then go ahead and post them together.)
* Do n... |
1d18c389-c17e-4670-ad8f-ec4decf53feb | trentmkelly/LessWrong-43k | LessWrong | Attend the 2025 Reproductive Frontiers Summit, June 10-12
There should be a community oriented towards the genomic emancipation of humanity. There isn't such a community, but there should be. It's a future worth investing our hope in—a future where parents are able to choose to give their future children the gift of g... |
11a664d7-59f1-4ddc-8f11-ee65b2dc18ea | StampyAI/alignment-research-dataset/arxiv | Arxiv | Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey
1 Introduction
---------------
Curricula are ubiquitous throughout early human development, formal education, and life-long learning all the way to adulthood. Whether learning to play a sport, or learning to become an expert in mathemati... |
3825ad87-1047-450e-a496-63c250f7e9b1 | trentmkelly/LessWrong-43k | LessWrong | Social status & testosterone
We’ve discussed signaling and status endlessly on LW; I think this is right up our vein: a 2011 review of research on the connections between famous male hormone testosterone and various forms of social interaction and especially social status, Eisenegger et al’s “The role of testosterone ... |
a6075956-98a6-4c9f-881f-16394d9747bc | StampyAI/alignment-research-dataset/special_docs | Other | The AI revolution and international politics _ Allan Dafoe _ EAG 2017 Boston-by Centre for Effective Altruism-video_id Zef-mIKjHAk-date 20170618
# The AI revolution and international politics (Allan Dafoe) - EA Forum
Artificial intelligence (AI) is rapidly improving. Superhuman AI in strategically relevant domains is... |
3efae9c8-f6df-4c33-ac06-52f78cfdc027 | trentmkelly/LessWrong-43k | LessWrong | Protectionism in One Country: How Industrial Policy Worked in Canada
TL;DR: the Canadian National Policy of 1879-1895 was a successful instance of deliberate infant industry protection (IIP). Selective targeting of industries with close domestic substitutes led to rapid output expansion, productivity growth, and price... |
ed80a351-2ece-4434-9d7d-5f5fcf9ec2d3 | trentmkelly/LessWrong-43k | LessWrong | Announcement: Writing Day Today (Thursday)
The MIRI Summer Fellows Program is having a writing day, where the participants are given a whole day to write whatever LessWrong / AI Alignment Forum posts that they like. This is to practice the skills needed for forum participation as a research strategy.
On an average da... |
e9110e9a-6807-469d-abc4-794ae3ae3b99 | trentmkelly/LessWrong-43k | LessWrong | Possibilities for keeping track of long threads
"Continue this thread" and "Comment score below threshold" could get a green glow if there are new comments behind the link.
An unthreaded version of Sort by New. |
aaca29ee-7320-48d7-b254-827f176f078e | trentmkelly/LessWrong-43k | LessWrong | Why is this utilitarian calculus wrong? Or is it?
Suppose that I value a widget at $30. Suppose that the widget costs the widget-manufacturer $20 to produce, but, due to monopoly power on their part, they can charge $100 per widget.
The economic calculus for this problem is as follows. $30 (widget valuation) - $100 (... |
2b61be1e-41bf-49ec-9903-2a6caa51f7ea | trentmkelly/LessWrong-43k | LessWrong | LLM-based Fact Checking for Popular Posts?
Hello,
I think having some LLM-based fact checking on LessWrong posts would be a valuable addition.
To some extent comments already serve this purpose, but LLMs can do this in a more automated and systematic way, side-stepping human blind spots and tendency to conserve ener... |
6deb7b07-a62c-4c2f-b1ec-0fceeef13db9 | trentmkelly/LessWrong-43k | LessWrong | Recommended Reading for Evolution?
I'll make this short and sweet.
I've been reading Dawkin's The Selfish Gene, and it's been really helpful filling in some of the gaps I have in my understanding of how evolution actually works.
The last biology class I took was in high school, and I don't think the mechanics of evo... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.